The Terraform migrate CLI lets you migrate Terraform state to HCP Terraform or Terraform Enterprise. The CLI allows you to migrate multiple files in a single operation.
In this tutorial, you will provision resources using multiple Terraform CLI workspaces, with state stored in AWS S3. You will then use the tf-migrate CLI tool to migrate those state files to HCP Terraform.
This tutorial assumes that you are familiar with the Terraform and HCP Terraform workflows. If you are new to Terraform, complete the Get Started collection first. If you are new to HCP Terraform, complete the HCP Terraform Get Started collection first.
For this tutorial, you will need:
Navigate to the template repository for this tutorial. Click the Use this template button and select Create a new repository. Choose a GitHub account to create the repository in and name the new repository learn-terraform-migrate. Leave the rest of the settings at their default values.
Clone your example repository, replacing USER with your own GitHub username.
$ git clone https://github.com/USER/learn-terraform-migrate
Change to the repository directory.
$ cd learn-terraform-migrate
The example configuration in this repository creates state files for you to use to test tf-migrate. Explore the example configuration.
.
├── example
│ ├── backend.config
│ ├── compute
│ │ ├── main.tf
│ │ ├── outputs.tf
│ │ └── variables.tf
│ └── storage
│ ├── main.tf
│ ├── outputs.tf
│ └── variables.tf
└── s3
├── main.tf
├── outputs.tf
└── terraform.tf
The configuration in each of the compute and storage directories generates random strings. Both of these root modules use the shared backend.config file to configure the S3 backend.
The s3 directory contains Terraform configuration to create an AWS S3 bucket. You will use this S3 bucket to store the state that you create.
Navigate to the s3 directory.
The configuration in this directory creates an AWS S3 bucket, enables versioning on the bucket, and creates the DynamoDB table. You will use this bucket to store your configuration's state.
Initialize the configuration.
Next, apply the configuration. Respond with yes when Terraform prompts you to confirm your changes.
$ terraform apply
## …
Do you want to perform these actions?
Terraform will perform the actions described above.
Only 'yes' will be accepted to approve.
Enter a value: yes
##...
Apply complete! Resources: 4 added, 0 changed, 0 destroyed.
Outputs:
backend_name = "learn-terraform-migrate-ee83uysoydtyn0io"
You will reference the backend_name output value later in this tutorial.
Next, navigate to the example directory to create the example state.
The configuration in this tutorial uses a shared configuration file to configure the S3 state backend. Refer to the partial configuration documentation to review how Terraform overlays backend configuration.
Open the backend.config file and set both the bucket and dynamodb_table values to the backend_name output from the previous step.
example/backend.config
bucket = "REPLACE_ME"
dynamodb_table = "REPLACE_ME"
region = "us-east-2"
Next, navigate to the storage directory.
Review the main.tf file.
example/storage/main.tf
terraform {
required_providers {
random = {
source = "hashicorp/random"
version = "3.6.3"
}
}
backend "s3" {
key = "learn-terraform-migrate/example/storage/terraform.tfstate"
}
}
resource "random_string" "this" {
count = var.string_count
length = var.string_length
}
This configuration uses the random provider to create multiple random_string resources. Next, review the variables.tf file.
example/storage/variables.tf
variable "string_count" {
type = number
description = "Number of strings to create"
default = 3
}
variable "string_length" {
type = number
description = "Length of each string"
default = 32
}
This configuration defines two variables, one named string_count and one named string_length.
Next, initialize your configuration and provide the path to your backend.config file.
$ terraform init --backend-config ../backend.config
Then apply your configuration. Respond with yes when Terraform asks you to confirm your changes.
$ terraform apply
##...
Do you want to perform these actions?
Terraform will perform the actions described above.
Only 'yes' will be accepted to approve.
Enter a value: yes
##...
Apply complete! Resources: 3 added, 0 changed, 0 destroyed.
Outputs:
strings = [
"m:xpV{*zb@?d<!O!RXJ<(@vv%p8iaj$t",
"JTW{GBf0425nOB2U+@TFu0p3_d{ZsmSy",
"xmr+:}S(kwJp$$yMtWj{TSu:1oQ0!h_O",
]
Next, navigate to the compute directory.
The configuration in this directory matches the configuration in the storage directory. It creates multiple random_string resources and defines the same two variables.
Initialize the configuration, passing the path to the backend.config file.
$ terraform init --backend-config ../backend.config
Next, apply the configuration. Respond with yes when Terraform asks you to confirm your changes.
$ terraform apply
##...
Do you want to perform these actions?
Terraform will perform the actions described above.
Only 'yes' will be accepted to approve.
Enter a value: yes
##...
Apply complete! Resources: 3 added, 0 changed, 0 destroyed.
Outputs:
strings = [
"]6+-EMu&q<nqxbDo&Y54u$:5Q4uG>Z]j",
"HH}a{]i=7yQ2jLMHe{hF:?uZu6s*?Btl",
"bf70A]szQx(PG<yAJ7(D#cW1FJ]yoyZm",
]
Next, create a new local workspace for the compute configuration to isolate the state data.
$ terraform workspace new testing
Created and switched to workspace "testing"!
You're now on a new, empty workspace. Workspaces isolate their state,
so if you run "terraform plan" Terraform will not see any existing state
for this configuration.
Run terraform apply again to create state in your new local workspace. Respond with yes when Terraform asks you to confirm your changes.
$ terraform apply
##...
Do you want to perform these actions in workspace "testing"?
Terraform will perform the actions described above.
Only 'yes' will be accepted to approve.
Enter a value: yes
##...
Apply complete! Resources: 3 added, 0 changed, 0 destroyed.
Outputs:
strings = [
":fI[%NL>Jrr>Y$kXOkvYXt6G-C1c)>rX",
"LH[jN<S49!@@y1qN#H&2vY1{urf&K%H:",
"Ck#UP1HkjlrR?VUwK#Cw]c)UJe[D9Bng",
]
Navigate back to the top-level directory of your repository.
You can use the aws CLI to see the three state files that you created in your S3 bucket.
$ aws s3 ls s3://$(terraform output -state s3/terraform.tfstate -raw backend_name) --recursive
2025-02-27 15:27:32 2566 env:/testing/learn-terraform-migrate/compute/frontend/terraform.tfstate
2025-02-27 15:27:26 2536 learn-terraform-migrate/compute/frontend/terraform.tfstate
2025-02-27 15:27:12 2536 learn-terraform-migrate/example/storage/terraform.tfstate
The tf-migrate CLI updates your state storage configuration and automatically opens a pull request to push the changes to version control. Migrating state to HCP Terraform using the tf-migrate CLI requires an HCP Terraform API token and a GitHub personal access token to create a pull request with the configuration changes.
By default, tf-migrate uses the same HCP Terraform API token that the Terraform CLI uses. If you have not previously logged in to HCP Terraform from the Terraform CLI, run the terraform login command. Respond yes to the prompt to confirm that you want to authenticate.
$ terraform login
Terraform will request an API token for app.terraform.io using your browser.
If login is successful, Terraform will store the token in plain text in
the following file for use by subsequent commands:
/Users/user/.terraform.d/credentials.tfrc.json
Do you want to proceed?
Only 'yes' will be accepted to confirm.
Enter a value: yes
Terraform automatically opens a browser window to the HCP Terraform login screen. Enter a token name in the web UI, or leave the default name, terraform login. Leave the default Expiration as "30 days".
Tip
If your browser does not open automatically, navigate to the URL output in your terminal or click here to go to the token creation UI.
Click Generate Token to generate the authentication token, then copy it to your clipboard.
In your terminal, paste the user token exactly once. Terraform will hide the token for security when you paste it into your terminal. Press Enter to complete the authentication process.
Create a GitHub personal access tokenFollow the instructions in GitHub's documentation to create a classic personal access token. To complete this tutorial, your personal access token must have the repo OAuth scope.
After you generate your personal access token, copy it to your clipboard. In your terminal, set the TF_GIT_PAT_TOKEN environment variable to the token.
$ export TF_GIT_PAT_TOKEN=
Using the tf-migrate CLI to migrate state requires two steps. The prepare step scans your current directory for Terraform configuration and state, generates Terraform configuration to perform the migration, and prompts to open a pull request against the code to capture the changes in version control. The execute step implements the changes.
Terraform migrate requires that you first commit all local changes to source control before you run the tf-migrate prepare command. Use the git commit command to commit your backend changes to source control.
$ git commit example/backend.config -m "Update S3 backend configuration"
Push your changes to your GitHub repository.
Next, run the tf-migrate prepare command with the skip-dir flag to skip the s3 configuration that contains your state bucket. Enter the name of the HCP Terraform organization that you would like to migrate your state to when prompted.
$ tf-migrate prepare -skip-dir="s3"
✓ Current working directory: /tmp/learn-terraform-migrate
✓ Environment readiness checks completed
✓ Found 1 HCP Terraform organization
┌────────────────────────────┐
│ Available Orgs │
├────────────────────────────┤
│ hashicorp-education │
└────────────────────────────┘
Enter the name of the HCP Terraform organization to migrate to: hashicorp-education
The tf-migrate prepare command generates Terraform configuration to migrate your state, including resource definitions for new HCP Terraform workspaces and projects, and updates your configuration to replace the backend block with the cloud block. When prompted to create a new local branch for this configuration, respond with yes.
✓ You have selected organization hashicorp-education for migration
✓ Found 3 directories with Terraform files
┌────────────────────────────┐
│ Terraform File Directories │
├────────────────────────────┤
│ example/compute │
│ example/storage │
│ s3 │
└────────────────────────────┘
The following 1 directories were skipped:
• s3
as they either have no backend, no supported backend to migrate or were excluded based on user-specified skip-dir arguments.
Create a local branch named hcp-migrate-main from the current branch main ... ?
Only 'yes' or 'no' will be accepted as input.
Type 'yes' to approve the step
Type 'no' to skip
Enter a value: yes
Terraform migrate then asks if you want to create a pull request in GitHub with your configuration changes. When prompted, respond with yes.
✓ Successfully created branch hcp-migrate-main
Do you want to open a pull request from hcp-migrate-main ... ?
Only 'yes' or 'no' will be accepted as input.
Type 'yes' to approve the step
Type 'no' to skip
Enter a value: yes
✓ Migration config generation completed
✓ Successfully updated .gitignore
Terraform migrate finishes the prepare step by generating the configuration to migrate your state and store it in the _hcp-migrate-configs directory. At this point, Terraform migrate only generates the configuration to perform the migration, it has not moved your state or modified your code.
Navigate to the _hcp-migrate-configs directory.
$ cd _hcp-migrate-configs
Review the configuration that tf-migrate generated.
.
├── main.tf
├── output.tf
├── variables.tf
└── workspace_variables_module
├── main.tf
├── output.tf
└── variable.tf
This configuration creates the HCP Terraform projects, workspaces, and workspace variables for your configuration. It also commits the changes tf-migrate made to your configuration, and opens the pull request in GitHub.
Open the variables.tf file. This file contains all of the information specific to your migration, including the workspaces and variables to migrate. If you need to customize your migration, you can modify this file as needed. For this tutorial, leave the values that Terraform migrate generated.
The tf-migrate execute command runs the Terraform configuration that the tf-migrate prepare command generated. This configuration changes updates the state storage location for your configuration, creates the pull request with those changes, migrates your state, and creates the workspace variables in HCP Terraform.
Navigate back to the top-level directory.
Next, run the tf-migrate execute command to run the migration.
$ tf-migrate execute
✓ Init command ran successfully
✓ Plan command ran successfully and changes are detected
✓ Apply command ran successfully
Apply complete! Resources: 20 added, 0 changed, 0 destroyed.
Migration Summary
┌───────────────────────────────┬───────┐
│ Metric │ Count │
├───────────────────────────────┼───────┤
│ Number of Projects Migrated │ 2 │
│ Number of Directories Skipped │ 1 │
│ Number of New Workspaces │ 3 │
│ Number of Variables Migrated │ 6 │
└───────────────────────────────┴───────┘
┌─────────────────────────────────────────────────────────────────────────────────────┐
│ Workspace URLs │
├─────────────────────────────────────────────────────────────────────────────────────┤
│ https://app.terraform.io/app/hashicorp-education/workspaces/example_compute_default │
│ https://app.terraform.io/app/hashicorp-education/workspaces/example_compute_testing │
│ https://app.terraform.io/app/hashicorp-education/workspaces/example_storage_default │
└─────────────────────────────────────────────────────────────────────────────────────┘
┌───────────────────────────────────────────────────────────────────────┐
│ Pull Request Link │
├───────────────────────────────────────────────────────────────────────┤
│ https://github.com/hashicorp-education/learn-terraform-migrate/pull/1 │
└───────────────────────────────────────────────────────────────────────┘
The tf-migrate command output displays a summary of its operation. In this example, tf-migrate created two projects, three workspaces, and migrated six variables. The summary also contains a link to all three workspaces and the GitHub pull request.
Navigate to your HCP Terraform organization. Terraform migrate created two projects, one named example_compute and example_storage. Notice that the Description field of each project records the repository that tf-migrate created it from.
Open the example_compute project and note that tf-migrate created two workspaces in this project, one named example_compute_default and the other named example_compute_testing. These match the two local CLI workspaces you created earlier.
Navigate to the example_compute_default workspace, then scroll down to the Resources table. This table contains a list of the three random_string resources that were in your local state. Next, click Variables in the left navigation menu. Notice that tf-migrate created the two workspace variables from your configuration.
Finally, navigate to the example_storage project and notice that tf-migrate only created one workspace in this project. Click the example_storage_default workspace. This workspace also contains three random_string resources from your local state, as well as the two variables that your configuration defined.
Open the repository you created earlier in GitHub and notice that you have one new pull request named "HCP Migration of Workspaces in USER/learn-terraform-migrate repository". You can merge this pull request to bring your repository's main branch up to date with the changes Terraform migrate made to your configuration.
Now that Terraform migrate has complete the migration, all future Terraform commands that you run against your configuration will run in HCP Terraform.
Delete the workspaces you created in this tutorial.
In HCP Terraform, navigate to the "example_storage" project and open the example_storage_default workspace. Navigate to Settings > Destruction and Deletion, then click the Force delete from HCP Terraform button. Since the only resources this workspace manages are random strings, you can safely delete the state without first running a destroy plan. Follow the prompts to delete the workspace.
Follow the same steps to delete the example_compute_default and example_compute_testing workspaces.
Next, delete the two projects that Terraform migrate created. Navigate to the example_storage project, click Settings, then click the Delete button. Follow the prompts to delete the project.
Follow the same steps to delete the example_compute project.
Finally, delete the AWS S3 bucket. In your terminal, navigate to the s3 directory.
Run terraform destroy to delete your resources. Respond with yes when Terraform prompts you to confirm.
$ terraform destroy
##...
Do you really want to destroy all resources?
Terraform will destroy all your managed infrastructure, as shown above.
There is no undo. Only 'yes' will be accepted to confirm.
Enter a value: yes
##...
Destroy complete! Resources: 4 destroyed.
In this tutorial, you used tf-migrate to automate migrating multiple state files to HCP Terraform, and reviewed how the CLI updates your configuration. You created three example state files using multiple directories and local workspaces. Then, you used Terraform migrate to migrate all of your state files to HCP Terraform automatically.
With your state in migrate, you can start using the feature that HCP Terraform provides to improve your infrastructure management lifecycle such as policy checks, drift detection, dynamic credentials, permission controls, and more.
To learn more about managing and migrating Terraform state, refer to the following resources:
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4